I'm logging. I don't understand '', Dave911. Try /msg logger_dev help
Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emcdevel/2009-11-03.txt
thanks... I forgot..;-)
mozmck, do you know how that compares to the same machine with 8.04?
(switched in here in case we want to get technical :) )
no, but it seems a little better than this machine with 9.04 I think it would run as high as 12us.
SMP/isolcpus on 9.04?
I made .deb files for the kernel and rtai.
no, same configuration.
no, I actually built what should be a very generic kernel.
same as the "normal" emc2 RTAI kernel, or same as you tried with 9.04? :)
ah, same configuration on 9.04 with as this. SMP, no isocpus (never got around to it)
kernel on that one was 188.8.131.52 vanilla with rtai 3.7.1 I think
from what I read there is no longer a problem with SMP configuration on single core machines with rtai 3.7.1, so I am planning to try this out on a number of different machines.
ok, so it's a relatively generic 32-bit SMP build?
yes. processor is i686 to get the TSC
so it should work on pentium pro on up.
I changed only what needed to be changed for the realtime stuff and left everything else at the default ubuntu settings.
I can send them in or put them up somewhere if anyone is interested...
the .debs that is
I can copy them to /experimental if you stick them somewhere I can access them
ok. I'll have to figure out somewhere to put them.
I wonder if I can give you an FTP URL to use
I've done FTP plenty...
I'm just not sure if anonymous write access is enabled anywhere
I think I can put them on my server here. Let me try right quick...
would you need the linux-source deb?
it wouldn't hurt, if you have it
that would give someone the exact config you started with
I think so. My link is only 512K so it will take a little bit to get them all.
when dealing with gpl software, you must be prepared to give corresponding source
uploading it in the first place is the best way to achieve that
they're all on the way to /experimental/karmic
yeah, that's true. My link is kind of slow and the source deb is ~60 meg :)
there's a badly formed index.html there now, so don't try going there yet
great! it would be good to see if they work for others.
hmmm. I'm heading for bed, so there will be a badly formed index.html until tomorrow :)
the transfer will likely take an hour or more
one thought I had is that I and maybe others can at least experiment with some of this before 10.04 and maybe get that out sooner.
it'll probably help a little
maybe there could be an SMP kernel by default in 10.04 if we can know it will work for most machines
but then again, we don't know what kernel version will be released with 10.04, or when/if an RTAI patch will be made available for it
"I can see where the MAXVEL slider might be handier than a whole case of bottled beer too."
to compile emc2 I had to make a change to hostmot2 as a couple of things have changed.
but I don't know if the change will work with the kernel in 8.04
I mean which branch are you talking about?
the one change should be fine, the other may not. they removed bus_id from the device struct in the new kernels, and you use dev_name and dev_set_name to access the device name now.
I don't know if those are in kernels before 2.6.27.
they also removed FIRMWARE_NAME_MAX but I just used an #ifndef and then #defined it, so that one should be fine
there is a macro to get the kernel version and compare it to a known value - maybe you can use this if you find that your changes break backward compatibility
ok, I was thinking of something like that. do you know what the macro name is?
hal_parport.c:214:#if LINUX_VERSION_CODE > KERNEL_VERSION(2,6,0)
thanks for working on fixing it! we need folks on the bleeding edge :-)
that's good :)
looks like there are about 4 other drivers that include linux/pci.h - do they all need to be updated or is it just a firmware loading thing?
nothing else gave an error, and this code is in the hm2_register function so it looks like it's just in his firmware loading stuff.
EMC: 03m-mcknight 07master * rf88ea06895db 10/src/hal/drivers/mesa-hostmot2/ (hostmot2.c hostmot2.h): make hostmot2 work on new kernels where FIRMWARE_NAME_MAX is removed and bus_id is removed from struct device
mozmck: thanks for fixing that. also it looks like unfortunately your editor (?) changed a few hundred lines you didn't intend to change. could you try to make it not do that, or use a different editor please? your real changes got buried in the noise in this diff.
uh-oh, how did it do that? it must have been gedit.
I wish I had a dollar for every time gedit has bitten one of us...
weird, I don't what it did. I'll look at the diff next time.
hm, looks like it (something?) removed trailing whitespace
maybe that's a git (mis)feature
it often whines about trailing whitespace - would not surprise me if the same person who made it whine also made it strip it somehow
or maybe it was eclipse? that's what I normally use and I haven't seen it do that before though.
oh I think I've heard that eclipse does stuff like that - there are settings somewhere - or was it kate??
my git is 184.108.40.206
well, no big deal, but yeah maybe review the git diff - would be nice if we knew what part of the process does it
I found it, there's a setting in eclipse for it that is checked.
that should fix it.
time for bed here see you later.
EMC: 03jthornton 07v2_3_branch * rfca9f84a11ff 10/docs/src/hal/drivers.lyx: Typo and markup fixes
apears to be a typo in the intergrater manual, page 196 table 31.1 2nd colum for classicladder
reads numPhysOuputs should be numPhysOutputs < t missing
EMC: 03jthornton 07v2_3_branch * r9a1da9e906ee 10/docs/src/ladder/classic_ladder.lyx: Fix Typo
jthornton: ../docs/src/Submakefile:392: *** Required image file ../docs/src/hal/images/parport-in.png does not exist. Stop.
2.3 or dev?
nm, I found it
EMC: 03jthornton 07v2_3_branch * rd810fdc07daa 10/docs/src/hal/images/ (parport-in.png parport-out.png): add images
jthornton: no, thank _you_
yea! the buildbot has all green lights
jthornton: by the way, don't document new probing yet -- I think it'll have to change further before it's ready
mozmck's 9.10 kernel/rtai .debs are available now in http://www.linuxcnc.org/experimental/Karmic/
jepler: just let me know when it is ready
cradek_ is now known as cradek
[14:49:59] <alex_joni> http://www.linuxcnc.org/component/option,com_kunena/Itemid,20/func,view/catid,27/id,1069/lang,en/#1079
ugh, saw that
I responded to the original and then thought better and deleted it - nice feature
* alex_joni looks at canned replies
"oh, how about you try mach?"
"how about you quit whining and go to the trouble of composing your thoughts and participating in a meaningful way?"
that assumes he has some way of composing his thoughts :P
I'm sure he does - he's just new and should have thought more before posting.
I did a bunch of stuff last night that involved switching between two tools - so nice to have a tool changer
I can imagine
* alex_joni runs home
alex_joni: how can I translate part of linuxcnc page?
(I mean preview and such)
micges: what language?
"preview" is silly, ignore it
I can set you up
do you have an account for linuxcnc.org ?
interesting .. :P
* alex_joni tries to guess it
joomla, not shell (AFAIK)
yes.. webpage account
same nick :)
alex_joni: ok and how can I add new language?
micges: working on it
Fatal error: Call to a member function getError() on a non-object in /home/.jared/cncman/www.linuxcnc.org/administrator/components/com_mambelfish/admin.mambelfish.php on line 132
alex_joni: may I do some russian translation too?
aystarik: yeah.. sure, if I get it going
aystarik: just translate the pages, and send them to me
or plain text, whatever suits you
no preference on your side?
micges: I managed to break it :/
will you fix it?;) I'll send you translations :)
yes I see, no others languages available
micges: it's mostly working now (the infrastructure at least)
but I lost all translations :/ trying to recover them now
micges: most of it is back :D
aystarik: ok.. now you can get my attention ;)
aystarik: it would help if you grap the html, and translate it. if not, simple text is ok too
(a bit more work for me)
micges: we'll continue this in a pm
yes, this is what I'm doing :)
I'll pm you my addy
Russian will require different code page -- 8859-1 does not have needed characters.
utf-8 should be ok.. we'll see
maybe do the front page only
to see if it works
yes, utf-8 is ok.
EMC: 03cradek 07master * r6decec1b49b4 10/src/emc/usr_intf/touchy/emc_interface.py: don't clear errors unnecessarily
does that look ok?
alex_joni: looks ok.
thumbnail screenshot is broken in firefox 3.5...
yes, it goes over voting on the right pane
english page is ok, but russian is not
and news are not translated.
looking at it now
sure, I just did the homepage
meaning the entry in the middle ;)
not the menus, not the polls, not the news
* alex_joni needs to do each one independently
aystarik: better now?
pic is ok, hyperlinks are black/not red as before...
yeah, saw that.. no idea why ;)
aystarik: one more attempt ;)
aystarik: I think it's easiest if I send you the html snippets, and you translate only those
(easier for me to integrate too)
looks ok now.
re: snippets -- ok.
I just sent you about.html
hope this way will be better
good night all
Guest718 is now known as skunkworks
aystarik: looks good: http://www.linuxcnc.org/content/view/11/10/lang,russian/
alex_joni: yep... Could you also recover header "Welcome" translation from index.html?
not sure what you mean
Ð.еÐ.Ñ.е пеÐ.аÐ.еваÑ.Ѧ на LinuxCNC.org ?
yes, this one... It should be UTF-8 encoding...
on the home page? or on the about page?
..... .......... .. .... ....... Enhanced Machine Controller(... EMC)
* alex_joni hates utf-8 + putty + irssi :D
Dobro pojalovati na sait proekta EMC ?
my cyrilic is not that good :D
dobro? na, banjo!
aystarik: where should I put it?
probably every russian page, the same place as for every other language...
it's only on the first page
I'm off to bed.. 2 am here :/
I'll send you files for the other translations
no, I mean the gray header words
ah, can't change those
that's part of the template, not content